Atrioventricular optimization in cardiac resynchronization therapy with quadripolar leads: should we optimize every pacing configuration including multi-point pacing?

This study aims to define an atrioventricular (AV) delay optimization method for cardiac resynchronization therapy (CRT) with a quadripolar left ventricular (LV) lead based on intrinsic conduction intervals. Heart failure patients with a left bundle branch block underwent CRT implantation with a quadripolar LV lead. Invasive LV pressure-volume loops were recorded during four biventricular and three multi-point pacing (MPP) settings, using four patient-specific paced AV delays. Haemodynamic response was defined as change in stroke work (Δ%SW) compared to intrinsic rhythm and was related to the following conduction intervals: right atrial pacing to right ventricular sensing interval (RAp-RVs), Q to LV sensing interval normalized to QRS duration (QLV/QRSd), PR-interval, and P-wave duration. In 44 patients, the largest Δ%SW (104 ± 76%) occurred at a paced AV delay of 128 ± 32 ms, at 47 ± 9% of RAp-RVs. Optimal AV delay of biventricular pacing (126 ± 26 ms) did not differ from MPP (126 ± 21 ms, P = 0.29). Intra-class correlation coefficient between optimal AV delays of different pacing configurations was 0.64 (0.45-0.78, P < 0.001). Although not statistically significant, Δ%SW at 50% of RAp-RVs (98 ± 74%) was closer to the maximal achievable Δ%SW increase than a fixed interval of 120 ms (96 ± 73%, P = 0.60). RAp-RVs, QLV/QRSd, PR interval, and P-wave duration were associated with the optimal AV delay in univariate analysis, but only RAp-RVs remained significantly associated in multivariate analysis (R = 0.69). The AV delay that provides highest haemodynamic response is similar for various LV pacing configurations and for MPP. An AV delay ∼50% of RAp-RVs creates an acute haemodynamic response close to the maximal patient-specific response.